Regeneration of insulin-producing beta-cells as well as islet transplantation is the most promising long-term solutions for Type 1 diabetes (T1D) treatment. A major obstacle to the success of this approach is the shortage of organ donors and the necessity for life-long immunosupression after allograft transplantation. Two recent advances, the success in reprogramming somatic cells into pluripotent ES cell-like IPS (induced pluripotent stem) cells and the success in differentiation of human ES cells into functional islet-like clusters (ILCs), have opened the possibility of generating patient-specific insulin-secreting ILCs through the reprogramming of somatic cells into iPS cells followed by guided differentiation. Indeed, in a proof-of-concept study, we have recently demonstrated that insulin-secreting cells can be generated from human skin fibroblasts using this approach. Despite this encouraging result, several obstacles have to be overcome before the value of this technique can be realized clinically. The major obstacles include: 1) safety concerns over the use of iPS cells that harbor viral vectors; 2) low efficiency of current protocols in generating functional beta-cells; 3) autoimmune destruction of the transplanted beta-cells. Given that the large iPS cell community is attacking the first problem and many immunology labs are trying to solve the third problem, the current proposal plans to use epigenetic, chemical, and ECM approaches to attack the second issue. The Specific Aims are: 1) Characterization of epigenetic features of fetal and mature human beta-cells; 2) Characterization of epigenetic features of definitive endoderm, pancreatic progenitors, and endocrine progenitors; 3) Identification of chemical molecules that promote beta-cell differentiation; 4) Characterization of the effect of extracellular matrix and other environmental factors on beta-cell differentiation;